Are All Taxes Equally Bad?
How Replacing Iowa’s Sales Tax Could Save Iowans
More Than $100 Million per Year
Harvey Lapan, GianCarlo Moschini, and Brad Caruth
Working Paper 02-WP 312
September 2002
Center for Agricultural and Rural Development
Iowa State University
Ames, Iowa 50011-1070
www.card.iastate.edu
Harvey E. Lapan is a University Professor, Department of Economics; GianCarlo Moschini is the
Pioneer Chair in Science and Technology Policy, Department of Economics and Center for
Agricultural and Rural Development; and Brad Caruth is a student, Department of Economics, all
at Iowa State University. We thank our colleagues Bruce Babcock and Dave Swenson for
comments and suggestions.
This publication is available online on the CARD website: www.card.iastate.edu. Permission is
granted to reproduce this information with appropriate attribution to the authors and the Center for
Agricultural and Rural Development, Iowa State University, Ames, Iowa 50011-1070.
For questions or comments about the contents of this paper, please contact Harvey Lapan, 283
Heady Hall, Iowa State University, Ames, IA 50011; Ph: 515-294-5917; Fax: 515-294-0221; Email: hlapan@iastate.edu; or GianCarlo Moschini, 583 Heady Hall, Iowa State University, Ames,
IA 50011; Ph: 515-294-5761; Fax: 515-294-6336; E-mail: moschini@iastate.edu.
Iowa State University does not discriminate on the basis of race, color, age, religion, national origin, sexual
orientation, sex, marital status, disability, or status as a U.S. Vietnam Era Veteran. Any persons having
inquiries concerning this may contact the Director of Affirmative Action, 318 Beardshear Hall, 515-294-7612.
Abstract
Under current U.S. law, taxpayers can deduct up to 100 percent of their state income
taxes from their adjusted gross income when calculating their federal income taxes. As a
result, Iowans currently pay approximately $251 million less to the federal government
than they would otherwise pay. There is, however, no equivalent stipulation allowing for
the deduction of state sales taxes. Consequently, by eliminating the sales tax and
replacing the lost revenue with an income-based tax, Iowans could save a substantial
amount of money on their federal tax returns without any change in revenue for the Iowa
government. Alternatively, by replacing the sales tax with an income-based tax, the State
of Iowa could increase its tax revenue without increasing the total tax burden on Iowans.
This analysis discusses four specific scenarios, with net benefits to Iowans ranging from
$106 million to $157 million per year.
Keywords: federal itemized deductions, income tax, sales tax, state budget, tax policy.
ARE ALL TAXES EQUALLY BAD? HOW REPLACING IOWA’S SALES
TAX COULD SAVE IOWANS MORE THAN $100 MILLION PER YEAR
Every tax ought to be so contrived as both to take out and to keep out of
the pockets of the people as little as possible over and above what it brings
into the public treasury of the state.
—Adam Smith, The Wealth of Nations, 1776
Introduction
Across the nation, state budgets are in disarray. In fiscal year 2002, state revenues
were 5.6 percent lower than forecasted, and an even bleaker scenario seems likely for
fiscal year 2003 (The Economist 2002). Iowa is no exception to this general situation, and
we are all familiar with the recent budget woes that have brought painful adjustments to a
number of critical state programs, especially education. Lawmaker and stakeholder
attention so far has taken a short-run perspective and has focused mainly on identifying
the size and nature of the cuts needed to balance the budget. There are reasons to believe,
however, that the budget problems currently facing Iowa and other states may not be only
a cyclical phenomenon related to the recent slowdown and recession in the general
economy but also may reflect deeper structural problems that require a long-run
reassessment of both spending and revenue options. In particular, in this paper we argue
that state governments should pay more attention to their choice of tax revenue
instruments, and that the State of Iowa could benefit substantially from the replacement
of the sales tax with an income tax.
The problem of taxation, to secure financial resources needed for public expenditures, has long been of interest to economists (Musgrave 1985). Issues that typically are
addressed relate to notions of “efficiency” and “equity.” Concerns about efficiency arise
because most taxes create distortions in the allocation of resources, such that they leave
taxpayers with a loss that is greater than the tax revenue collected by the government.1
Considerations of equity, on the other hand, are germane because taxpayers are
heterogeneous on many dimensions, and their welfare may be affected differently by a
2 / Lapan, Moschini, and Caruth
given tax. Equity in taxation is often understood to mean that individuals should be taxed
according to their ability to pay. But equity considerations are also at the root of the
principle that taxes should be levied on individuals according to the benefits they receive
(from the public expenditures financed by taxes). Unfortunately, equity and efficiency
considerations in taxation often lead to conflicting prescriptions. Striking a balance
between efficiency and equity leads to the problem of optimal taxation, which is widely
studied in economics.2
Given the difficulties of devising and implementing outright an optimal tax structure
in a real-world economy, the tax problem also has been formulated in terms of gradual
changes or policy reform (Feldstein 1976). In the spirit of this approach, this paper
articulates a specific tax policy reform that offers obvious advantages to the State of
Iowa. Specifically, the policy change advocated is to replace the sales tax with an
income-based tax. For a general economy, a prescription such as this need not have
theoretical validity, and indeed it seems at odds with the general trend of most developed
economies, as most nations have adopted a mix of tax instruments that involve both
direct taxation (e.g., income taxes and property taxes) and indirect taxation (e.g., sales
taxes and value-added taxes). But, in determining what is optimal for the State of Iowa, it
is crucial to account properly for all the constraints set forth by the federal government,
and it turns out that a feature of U.S. federal tax law provides a strong incentive for
individual states to favor a tax system based on income and/or property taxes.
Under current U.S. law, taxpayers can deduct up to 100 percent of their state income
(and property) taxes from their adjusted gross income when calculating their federal
income taxes.3 There is, however, no equivalent stipulation allowing for the deduction of
state sales taxes. Consequently, by eliminating or reducing the sales tax and replacing the
lost revenue with an income tax increase, Iowans could save a substantial amount of
money on their federal tax returns without any change in revenue for the Iowa
government. Alternatively, the Iowa government could increase revenue without
increasing the total tax burden on Iowans. As an illustration, we can point out that Iowans
at present already derive sizeable benefits from the federal deductibility provision:
Iowans currently pay approximately $251.3 million less to the federal government than
they would be required to pay if the deduction of state income taxes were not allowed.
Are All Taxes Equally Bad? / 3
In what follows, we articulate and document our proposal to replace the current sales
tax with an equivalent income tax. Whereas our proposal produces unambiguous and
sizeable gains to Iowans, any practical implementation of the idea underlying our
proposal will inevitably have some distributional consequences. Thus, we cannot escape
the trade-off between efficiency and equity of taxation, discussed earlier. Indeed, there
are many ways to implement the tax change discussed here, each with different impacts
on the distribution of the tax burden. Here, we concentrate on illustrating the general
principle, and we discuss perhaps the simplest implementation of the tax change, that is,
one based on a supplemental ad hoc flat income tax replacing the sales tax (while leaving
unchanged the current income tax structure of the State of Iowa). We estimate that such a
change to the Iowa tax system would produce approximately $157 million per year in
benefits to Iowans. Furthermore, these gains would be possible without a significant shift
in the distribution of the tax burden.
Background on the Iowa State Tax System
Iowa, like many states, relies on a variety of taxes and fees for its revenue, but the
majority of state revenue is obtained through the individual income tax and the retail
sales tax. The revenue generated from Iowa taxes and the relative importance of these
taxes is shown in Figure 1.
Individual Income Tax
The individual income tax is based essentially on federal adjusted gross income
(AGI) calculations. As taxable income increases through Iowa’s nine statutory income
brackets, the marginal tax rate increases from 0.36 percent to 8.98 percent as shown in
Table 1. The income brackets are indexed for inflation and are adjusted annually in order
to ensure that tax rates reflect real income. Figure 2 shows that, although Iowa’s income
tax schedule appears to be relatively progressive,4 the progressivity is lessened by the
federal income tax deduction. Even though federal tax laws allow all state income taxes
to be deductible, only Iowa, Louisiana, and Alabama allow all federal income taxes to be
deductible on state tax returns. The deductibility of federal income taxes on Iowa tax
returns has little impact on low-income taxpayers but it significantly reduces the tax
burden, and effective tax rate, of higher income taxpayers. The result is that low-income
4 / Lapan, Moschini, and Caruth
Source: Iowa Department of Revenue and Finance, November 2001.
FIGURE 1. Iowa state tax revenue by source, fiscal year 2001 (millions of dollars)
TABLE 1. Iowa’s statutory income tax rates
Taxable Income ($)
Marginal Tax Rate (%)
0 to 1,162
1,162 to 2,324
2,324 to 4,648
4,648 to 10,458
10,458 to 17,430
17,430 to 23,240
23,240 to 34,860
34,860 to 52,290
52,290 and over
0.36
0.72
2.43
4.50
6.12
6.48
6.80
7.92
8.98
Source: Iowa Department of Revenue and Finance, November 2001.
tax liabilities remain essentially unchanged while high-income earners are able to reduce
their taxable income.
Retail Sales Tax
The retail sales tax in Iowa is a tax of 5 percent on the sale of tangible goods and some
specifically designated services.5 In order to lessen its regressivity, some items are
exempted from the state sales tax (for example, food and medical expenses fall in this
Are All Taxes Equally Bad? / 5
Source: Authors’ calculations. See the Appendix for details.
FIGURE 2. Change in average state income tax rate due to federal income tax
deduction
category). Iowa law also allows for a maximum of 2 percent in additional local option
sales taxes. One percent of the local option sales tax is designated as the regular local
option tax, while the other 1 percent is designated as the school infrastructure local option
tax. The regular local option tax can be imposed in cities, rural areas, or countywide and
requires passage by general election. The school infrastructure local option tax must be
countywide and must be repealed within ten years after it is imposed.6 It also requires
passage by general election.
Distribution of Income and the Tax Burden
Figure 3 shows the distribution of income for Iowa taxpayers.7 Each bar shows the
percentage of taxpayers whose income falls in the range specified. The State of Iowa has
less income inequality than do most other states. This is due largely to the relatively low
number of very high-income people in the state, which decreases the dispersion of the
income distribution. The vast majority of Iowa taxpayers have an AGI of less than $70,000.
6 / Lapan, Moschini, and Caruth
Source: Authors’ calculations based on Iowa Department of Revenue and Finance, July 2001.
FIGURE 3. Income distribution in Iowa, tax year 1999
The incidences of the retail sales tax and the individual income tax, as a percentage
of AGI, are shown in Figure 4. It is interesting to note that the combined sales and
income taxes form a nearly proportional state tax system. Other state taxes, however, tend
to make the overall net tax burden more regressive.8
Interstate and Historical Comparisons
Table 2 shows that Iowa and neighboring states rely on a mix of income and sales
taxes to raise revenue. Iowa’s income tax schedule may appear to be more progressive
than the other states, but it is important to note that the higher tax rates for higher income
taxpayers in Iowa are partly offset by the federal income tax deductibility.9
Over time, however, the sales tax and the income tax have undergone several
changes. The Iowa state sales tax was first introduced in 1934 at a rate of 2 percent. The
most recent sales tax increase occurred in 1992 and raised the rate from 4 percent to 5
Are All Taxes Equally Bad? / 7
Source: Authors’ calculations. See the Appendix for details.
FIGURE 4. Sales and income tax incidences for Iowa, fiscal year 1999
TABLE 2. The income and sales tax rates of Iowa and its neighboring states
Income Tax Rate
Sales Tax
State
(lowest-highest)
Rate
Iowa
Illinois
Minnesota
Missouri
Nebraska
South Dakota
Wisconsin
0.36% to 8.98%
Flat 3%
5.35% to 7.85%
1.5% to 6.0%
2.51% to 6.68%
None
4.6% to 6.75%
5%
6.25%
6.5%
4.225%
5%
4%
5%
Source: Rates compiled by authors from data available on web pages of states’ departments of revenue.
percent. The income tax was first introduced in 1934 and ranged from 1 percent to 5
percent. The most recent change occurred for the 1998 tax year when all marginal rates
were cut by 10 percent of their value. The effect of this income tax cut on the trend of the
state’s revenue is evident in Figure 5, which shows Iowa’s gross tax collections for the
last 16 years. Figure 6 shows the percentage of all state taxes accounted for by each tax.10
8 / Lapan, Moschini, and Caruth
Source: Compiled by the authors from Iowa Department of Revenue and Finance Annual Reports (2000;
2001a,b)
FIGURE 5. Gross tax collections in Iowa, fiscal years 1986-2001
Source: Compiled by the authors from Iowa Department of Revenue and Finance Annual Reports (2000;
2001a,b)
FIGURE 6. Selected Iowa taxes as a percentage of gross state tax revenue, fiscal years
1986-2001
Are All Taxes Equally Bad? / 9
The nationwide trend for states has been an increasing reliance on sales taxes and a
decreasing reliance on income taxes.11 Iowa’s most recent sales tax increase and income
tax decrease fit this trend.
The Value of Replacing the Sales Tax with an Income-Based Tax
At present, Iowans pay both sales taxes and state income taxes. The basic reason to
consider the tax change analyzed in this paper has to do with the federal deductibility of
state income taxes (a deductibility that does not extend to sales taxes).12 To begin our
analysis, we first look at the value of the federal deductibility of state income taxes. It
turns out that Iowans currently pay approximately $251.3 million less to the federal
government annually because federal law allows taxpayers to deduct state income tax
payments from their federal taxable income.13 The distribution of these savings is shown
in Table 3.
On average, these savings translate to about $138 per Iowa tax return per year, but
high-income taxpayers realize the majority of the benefits. There are two reasons why
these benefits accrue mainly to high-income taxpayers: (i) because these taxpayers pay
high Iowa taxes and hence the deduction reduces their federal taxable income more; and
(ii) because, since their income is taxed at a higher rate, a given deduction is worth more
TABLE 3. Current tax savings by Iowans per year, due to federal deductibility of
state income tax (2002 dollars)
Savings as a
Total
Savings per
Percentage of AGI
(%)
AGI Bracket
Savings ($)
Return ($)
Less than 10,000
10,000 – 20,000
20,000 – 30,000
30,000 – 40,000
40,000 – 50,000
50,000 – 70,000
70,000 and over
All returns
43,347
1,791,796
6,925,659
15,154,473
11,312,143
23,277,589
192,837,014
251,342,020
Source: Authors’ calculations. See the Appendix for details.
0
8
38
62
83
184
1,994
138
0.00
0.03
0.08
0.19
0.20
0.44
1.28
0.49
10 / Lapan, Moschini, and Caruth
to them than it is to low-income taxpayers. Although taxpayers must itemize their federal
tax returns in order to benefit from the deductibility of state income taxes, no taxpayer is
hurt by the availability of this federal deduction provision.
The fact that the federal deductibility of state income (and property) taxes is quite
valuable to Iowans suggests that perhaps Iowa is not optimally exploiting this federal
deductibility provision. In fact, sizeable efficiency gains are possible by retooling Iowa’s
tax collection mechanisms, away from sales taxes and toward an income-based tax. What
remains to be determined is the actual monetary value of this policy change.
The Revenue-Neutral Lump-Sum Equivalent Tax
To see how we can put a dollar value on the tax policy change analyzed in this study,
suppose that we can observe the sales taxes paid by each taxpayer over the course of the
tax year. Let such an amount for the ith taxpayer be labeled as Ti . Now suppose that all
sales taxes are eliminated, and that a taxpayer’s state income liability is increased by a
lump-sum amount equal to Ti (over and above the current income tax, as per the existing
income tax structure). What would be the consequences of such a change? Clearly, the
government would collect exactly the same revenue from this taxpayer. As for the
welfare of the taxpayer, note that he or she can still afford the same consumption bundle
chosen under the sales tax system, and so from that perspective the taxpayer is equally
well off.14 But if the taxpayer is itemizing on his or her federal return, and if the lump
sum tax Ti were now deductible from the income subject to federal taxes, this change
would reduce the individual’s federal tax liabilities and result in substantial savings to the
taxpayer,15 say Si ≥ 0 .
If we could compute the amount Si saved by each taxpayer, then summing such
nonnegative amounts over all taxpayers would provide us with a monetary estimate of the
value of the policy change that replaces the sales tax by an individual-specific lump-sum
tax (sometimes called a “poll tax.”). Of course, we do not have enough information to
actually compute this amount for each and every taxpayer. But the conceptual experiment
carried out suggests a useful approximation based on a “representative taxpayer” for each
of the income classes that we identify (see the Appendix for details). Figure 7 shows the
sales tax paid per return, and Table 4 shows the amount saved by itemizing under this
Are All Taxes Equally Bad? / 11
Source: Authors’ calculations. See the Appendix for details.
FIGURE 7. Estimated sales tax per return in Iowa (2002 dollars)
TABLE 4. Estimated distribution of savings, per year, from itemizing the revenueneutral lump-sum tax (2002 dollars)
Savings as a
Total
Savings
Percentage of AGI
AGI Bracket
Savings ($)
Per Return ($)
(%)
Less than 10,000
10,000 to 20,000
20,000 to 30,000
30,000 to 40,000
40,000 to 50,000
50,000 to 70,000
70,000 and over
All Returns
604,748
3,350,695
6,588,183
10,718,909
6,361,156
11,489,229
67,197,153
106,310,073
Source: Authors’ calculations. See the Appendix for details.
1
8
18
44
47
113
554
58
0.03
0.05
0.07
0.12
0.10
0.20
0.42
0.19
12 / Lapan, Moschini, and Caruth
scenario. High-income taxpayers obtain the largest portion of the savings in this
counterfactual simulation because they currently pay the greatest percentage of the sales
taxes and have a higher marginal federal income tax rate. High-income taxpayers also are
more likely to itemize their federal income taxes. Table 4 shows the net outcome of the
lump-sum equivalent tax: Iowa taxpayers would gain $106.3 million dollars annually,
when measured in 2002 dollars.16 Table 4 also shows the calculated distribution of
savings by income bracket. Although no taxpayers would lose money, the distributional
impact of this scenario would be that of a regressive tax cut.
The Taxpayer-Neutral Lump-Sum Equivalent Tax
An alternative way of valuing the policy change is to suppose that the current sales
tax is replaced by an individual-specific lump-sum tax, the level of which is set so that
the State of Iowa collects the revenue gains arising from the policy change. Thus, instead
of replacing the sales tax by a lump-sum tax Ti equal to the sales tax paid by the ith
individual, the state replaces the sales tax by a lump-sum tax Tˆi ≥ Ti , which has the
property of leaving the taxpayer with the same after-tax disposable income as the
individual had under the sales tax system. In other words, the state government
anticipates that taxpayers who itemize would get a net benefit from the policy change and
the state adjusts the lump-sum tax accordingly. Summing the difference (Tˆi − Ti ) over all
taxpayers again gives us a monetary value for the policy change in terms of increased tax
revenue while leaving each and every taxpayer as well off as under the sales tax system.
As earlier, we cannot compute the amounts (Tˆi − Ti ) for each and every taxpayer, but
we can get a useful approximation based on the data for the representative taxpayers for
each income class. The net result for this scenario is that the Iowa government would
raise an additional $133.9 million annually (in 2002 dollars). Note that the estimated
“value” of the tax change here is higher than in the preceding scenario. To understand
why that is the case, suppose that Tˆi is initially set such that the state captures the entire
efficiency gain estimated in the preceding scenario ($106.3 million, in the aggregate). But
itemizing taxpayers still would have some gain, as explained in endnote 15. The state can
Are All Taxes Equally Bad? / 13
then internalize this further second-order gain, leading to the higher aggregate estimated
value of $133.9 million.
We need to re-emphasize that the calculations just discussed should be interpreted as
providing a benchmark estimate of the net benefits that would flow from replacing the
current sales tax with a lump-sum income tax. Of course, it is apparent that an individualspecific lump-sum tax is not a feasible alternative. But the idea of an individual-specific
lump-sum tax is useful for getting an estimate that is “conservative,” and immune, as
much as possible, from the ancillary distributional effects that pertain to the more realistic
mechanism that we shall describe in what follows. What we find is that there are
significant monetary gains that can be obtained by moving away from the sales tax, two
estimates for which are $106.3 million and $133.9 million, per year.17
Implementing the Tax Policy Change: The “One Line” Solution
The case of an individual lump-sum tax where the government keeps all the benefits
of the tax change, discussed in the previous section, is the only one where there are no
distributional impacts. But such an individual lump-sum tax is not feasible, and feasible
mechanisms for the proposed tax change are bound to have some distributional
consequences. It should be noted, however, that this is not necessarily bad from a policy
perspective. Because the potential tax savings that we have uncovered are the result of
itemizing federal income tax returns, a more progressive income tax increase may allow
for greater gains. (For a given income tax increase, a high-income earner will be more
likely to itemize and will be able to save more from itemizing.)
Among the many feasible ways to replace the sales tax with an income-based tax, in
this section we illustrate two alternative approaches. What these two approaches have in
common is simplicity: they would require one additional line in the Iowa 1040 form, a
“sales tax replacement” line with the amount of the additional (sales-tax replacement) tax
proportional to taxable income. The first alternative is based on the analysis of the
previous section, and would replace the current sales tax with an income-based
replacement tax whose rate changes according to the income class of the taxpayer. We
refer to this scenario as the “differentiated rates” solution. The second alternative is
simpler, and would replace the current sales tax with an income-based tax whose rate is
14 / Lapan, Moschini, and Caruth
the same for everyone. We refer to this scenario as the “flat tax” solution. For each of the
two approaches, we report two sets of estimates: one where the state’s tax revenue
remains the same before and after the tax policy change (the “revenue-neutral” scenario),
and one where the average tax burden of the taxpayer remains the same (the “taxpayerneutral” scenario). In the first of these two polar cases, the taxpayers get all the benefits
because of the efficiency gains arising from the tax policy change, whereas in the second
case it is the state that captures all the benefits.
The “Differentiated Rates” Solution
Previously, we computed the “value” of replacing the sales tax with an income-based
tax by assuming that there is a representative taxpayer for each income class that we
identified. The analysis of that section can be readily adapted to produce income tax rates
that would correspond to the estimated lump-sum taxes. Consider first the “revenueneutral” scenario. The task here is to ensure that the State of Iowa collects the same total
tax revenue. Having determined the lump-sum tax (the tax for each taxpayer) within each
income class, as in the previous section, and then dividing that amount by the average
taxable income of that class produces the implicit tax rate for that class. The procedure
for the “taxpayer-neutral” scenario is essentially the same (what is different is the
estimated level of the lump-sum tax for each class, as discussed in the previous section).
The results of this analysis are reported in Table 5.
For the revenue-neutral scenario, the tax rates required to replace the sales tax
decline uniformly as the taxable income increases. This is because taxpayers with lower
income levels spend a higher proportion of their income on goods and services that are
subject to the sales tax. Indeed, that is the reason why the sales tax is commonly
perceived to be “regressive.” The pattern is similar for the taxpayer-neutral scenario. But
because here the task is to hold constant the total tax burden of the representative
taxpayer for each class, the interplay of federal and state marginal income tax rates is
important. It turns out that the tax rates required to replace the sales tax for this scenario
decline as taxable income increases, up to the AGI level of $40,000, after which the rate
is essentially flat (and approximately equal to 3 percent).
Are All Taxes Equally Bad? / 15
TABLE 5. Estimated distribution of tax rates to replace the sales tax revenue-neutral
and taxpayer-neutral scenarios
AGI Bracket ($)
Revenue-Neutral Tax (%)
Taxpayer-Neutral Tax (%)
Less than 10,000
11.30
11.34
10,000 to 20,000
20,000 to 30,000
30,000 to 40000
40,000 to 50,000
50,000 to 70,000
70,000 and over
4.99
3.85
3.32
2.87
2.77
2.26
5.06
3.94
3.49
3.01
3.07
3.07
Source: Authors’ calculations. See the Appendix for details.
In some sense, the solution illustrated in Table 5 is the one with the most limited
distributional consequences. By construction, the rates in Table 5 would ensure that no
“representative taxpayer” is negatively affected by replacing the sales tax with an
income-based tax. Of course, this procedure does not account for the heterogeneity of
taxpayers within each AGI class. So, although taxpayers are not negatively affected on
average, it is possible that some may lose while others experience a (larger) gain. Perhaps
more important, the differentiated tax rates in Table 5 may not be very appealing to
policymakers, as they may be perceived as being discriminatory toward lower income
taxpayers (although, in fact, these rates essentially mimic the effects of the current sales
tax). For that reason, in the next section we consider the single-rate (flat tax) alternative.
The “Flat Tax” Solution
With this alternative, the sales tax is replaced with an income tax calculated by
applying a flat rate (a rate which is the same for all taxpayers) to taxable income.
Consider first the “revenue-neutral” scenario. Our calculations (see the Appendix) show
that, in order to maintain the same total tax revenue, the current 5 percent tax on sales
would have to be replaced with a 3.50 percent tax on income. The new marginal income
tax rates that would result are shown in Table 6.
It is important to emphasize again that, in the proposal of this study, the current
income tax structure is taken as given. Thus, the final rates comprise both the old
marginal tax rates and sales tax replacement. Because the sales tax would be replaced
16 / Lapan, Moschini, and Caruth
TABLE 6. The effect of the sales-tax-replacement equivalent flat tax on the statutory
marginal tax rates
Current
Flat Tax
Total
Taxable Income Marginal Tax Rate
Equivalent
Marginal Tax Rate
($)
(%)
to Sales Tax (%)
(%)
0 to 1,162
0.36
3.50
3.86
1,162 to 2,324
0.72
3.50
4.22
2,324 to 4,648
2.43
3.50
5.93
4,648 to 10,458
4.50
3.50
8.00
10,458 to 17,430
6.12
3.50
9.62
17,430 to 23,240
6.48
3.50
9.98
23,240 to 34,860
6.80
3.50
10.30
34,860 to 52,290
7.92
3.50
11.42
52,290 and over
8.98
3.50
12.48
Source: The first column is from the Iowa Department of Revenue and Finance (November 2001), whereas
the last two columns arise from the authors’ calculations (see the Appendix for details).
with an income tax, this change would tend to benefit people who spend a larger portion
of their income. Consequently, an added distributional effect occurs in this scenario
because low-income taxpayers tend to spend a larger percentage of their income than do
high-income taxpayers. Figure 8 shows the average tax burden (from income and sales
tax only) before and after the sales tax replacement. Although the net benefit to Iowa
taxpayers is $141.4 million dollars annually (in 2002 dollars), Figure 8 shows that
replacing the sales tax with a flat tax also has a progressive distributional impact in which
low-income taxpayers have a lower average tax rate, whereas high-income taxpayers
actually have a higher average tax rate.
Although the high-income taxpayers obtain the greatest benefit from itemizing the
deductions on their federal returns, this does not fully counteract the increased average
tax rate. Conversely, the low-income taxpayers obtain the least benefit from itemizing the
deductions on their federal returns, but the gains from the sales tax elimination more than
counteract the losses from the income tax increase. The result is that the average member
of the lowest income group gains $274 on his or her tax return whereas the average
member of the highest income group loses $380. Table 7 shows the calculated
distribution of savings by income bracket.
Are All Taxes Equally Bad? / 17
Source: Authors’ calculations. See the Appendix for details.
FIGURE 8. Average tax rates before and after sales tax replacement by flat tax
equivalent
TABLE 7. Distribution of savings due to itemizing the revenue-neutral flat tax
Savings as a
Savings
Percentage of AGI
AGI Bracket ($)
Total Savings ($)
Per Return ($)
(%)
Less than 10,000
121,832,006
274
5.12
10,000 to 20,000
75,198,360
182
1.17
20,000 to 30,000
31,099,388
85
0.33
30,000 to 40,000
-1,312,961
-5
-0.02
40,000 to 50,000
-22,275,043
-164
-0.36
50,000 to 70,000
-17,050,760
-168
-0.30
70,000 and over
-46,052,774
-380
-0.28
All returns
141,438,217
77
0.26
Source: Authors’ calculations. See the Appendix for details.
18 / Lapan, Moschini, and Caruth
The flat tax equivalent replacement of the sales tax can also be set such that the state
captures the efficiency gains made possible by the federal deduction provision. Thus, in
this scenario, the total tax burden on Iowans remains the same and the state receives the
additional revenue that is generated. For this case, we estimate that the 5 percent sales tax
would have to be replaced with a 3.89 percent income tax (see the Appendix for detail).
As in the previous scenario, replacing the sales tax (which tends to be regressive)
with a proportional income-based tax creates a distributional effect in which the poor
benefit more than do wealthier taxpayers. Overall, the total tax bill for Iowans will be
unchanged, but taxpayers will be paying more money to the state government and less
money to the federal government. The net benefit to the Iowa state government is
estimated to be $157 million annually (in 2002 dollars). Table 8 shows the calculated
distribution of savings by income bracket. Again, when the state captures the efficiency
gains of the tax reform, the estimated value of the policy change is greater than the
corresponding value that applies when the taxpayers capture the efficiency gains.18
Ancillary Reasons to Favor the Proposed Tax Change
Despite the obvious benefits of replacing the sales tax with an income tax, any
changes in Iowa tax law are likely to encounter some opposition from Iowa’s citizens and
legislators. Novelties are seldom welcome in this arena, as illustrated by the maxim “an
old tax is a good tax.” But the question naturally arises: whom or what is the old tax good
TABLE 8. Distribution of savings due to itemizing the taxpayer neutral flat tax
Savings as a
Savings
Percentage of AGI
Per Return ($)
(%)
AGI Bracket ($)
Total Savings ($)
Less than 10,000
115,913,619
260
4.87
10,000 to 20,000
56,718,327
137
0.88
20,000 to 30,000
3,529,411
10
0.04
30,000 to 40,000
-26,002,574
-106
-0.30
40,000 to 50,000
-39,232,454
-289
-0.64
50,000 to 70,000
-31,720,380
-313
-0.56
70,000 and over
-79,205,949
-654
-0.49
Total
0
0.00
0.00
Source: Authors’ calculations. See the Appendix for details.
Are All Taxes Equally Bad? / 19
for? In this study we have endeavored to show that the “old” sales tax may, in fact, not be
a good tax, not for Iowans at least. In support of the proposed tax change we have
provided estimates of substantial statewide efficiency gains. But there are other ancillary
advantages that should be noted when assessing the proposal put forth in this study.
First, elimination of sales taxes may have some positive impact on local retailing and
have a general pro-business effect, as the reduction in retail prices would make locally
sold goods and services more attractive to Iowans and out-of-state residents.
Second, the sales tax is becoming a less and less effective instrument for state tax
collection, because of the rise of Internet commerce and because of an additional trend
toward an eroded sales tax base.19 Bruce and Fox (2001) estimate the total loss in U.S.
states and local tax revenue due to e-commerce in year 2001 to be $ 13.3 billion, and,
perhaps more important, they forecast that this loss will increase fourfold by 2011. But
regardless of Internet commerce, there has been a trend that is eroding the sales tax base
nationwide, namely that whereas sales taxes predominantly target goods, a larger and
increasing portion of consumer expenditures is allocated to services.
Third, eliminating the sales tax would reduce the cost of collecting tax revenue (for
both businesses and the state government) and the cost of monitoring tax compliance.
And, because the sales tax would be replaced with an income-based tax within an already
existing income-tax collection system, essentially no new burden would be required at
the collection and monitoring stage.
One potential shortcoming of the income-based tax that we are proposing is that
income taxes are known to distort the labor market. Specifically, an increased income tax
may adversely affect people’s willingness to work, resulting in a lower supply of labor
and an increased consumption of “leisure,” the quintessential untaxed good. The
empirical extent of this distortion, and its applicability to the case studied here, is unclear.
Finally, a recognized useful feature of the current sales tax system is that it
provides for “local options” to raise tax revenue. Whereas this flexibility is definitely
desirable, it is clear that such local options could also be easily implemented within an
income-based equivalent tax. Indeed, a similar option is already present in the current
Iowa income tax: the school district surtax is calculated as a percentage of the
taxpayer’s Iowa income tax liability.
20 / Lapan, Moschini, and Caruth
Conclusion
In this study we have proposed a tax policy change that holds the potential of
considerable benefits to Iowa taxpayers. The reason is that Internal Revenue Service rules
allow state income taxes to be deducted when computing taxable income from the federal
perspective, and an equivalent deductibility is not present for sales taxes. Because of that,
sales taxes constitute an “inferior” technology for raising state tax revenue. Thus,
replacing the sales tax with an income-based tax could improve the efficiency of Iowa tax
revenue collection in a sizeable way. For example, if the state were to adopt the taxpayerneutral flat tax, which is equivalent to the existing sales tax, we estimated that the State of
Iowa would increase its revenue by $157 million per year, without increasing the overall
tax burden on Iowans.
The magnitudes of the benefits of the policy change presented here are, of course
estimates. The methodology employed to compute such estimates could be refined, and a
more complete database could be developed in an attempt to get more accurate estimates.
Alternative experiments with different computational procedures have shown that the
estimated gains do not change significantly. Overall, our assumptions and procedure
actually provide somewhat conservative estimates, and a more complete methodology is
likely to yield slightly larger benefits for the scenarios analyzed here.
The recent recession and revenue shortfall have forced the governor and the state
legislators to cut or limit a variety of worthwhile services and programs. Replacing the
sales tax with an income-based tax, as proposed in this paper, could contribute positively
to mitigating the current financial problems confronting the State of Iowa. To illustrate
further the magnitude of our estimated efficiency gains, we note that $157 million dollars
in additional revenue could translate to annual salaries for more than 5,600 new teachers;
or, it could translate to more than 1.7 times the state funding of the University of
Northern Iowa; or, it could provide approximately 1.4 times the state’s current spending
on substance abuse programs.20 The federal government assumes the only major loss
under the proposed policy change. Replacing the state sales tax with an income-based tax
would seem to be unquestionably a positive change for Iowans.
Appendix
Methodology and Sources of Data
The analyses that we carried out relied primarily on tax data from the Iowa
Department of Revenue and Finance and from the Internal Revenue Service. We also
made use of the Consumer Expenditure Survey and the Consumer Price Index, both of
which are published by the Bureau of Labor Statistics.
Table 9 shows information about federal income tax returns that were filed by
Iowans. The data were used for the purpose of calculating the average marginal tax rate
for each income group and for calculating the percentage of taxpayers in each group that
itemizes their federal deductions. In order to calculate the average marginal tax rate, the
average taxable income per return was computed for each income bracket. After this was
completed, the portion of tax returns that were joint returns, single returns, and head of
household returns were calculated. For each type of return, the average taxable income in
each bracket was matched with the statutory marginal tax rate that applies. The average
marginal tax rate was then calculated by averaging the statutory marginal tax rates of the
three types of returns. The percentage of taxpayers that itemized their deductions was
found by dividing the number of itemized returns by the total returns count.
Tables 10 and 11 show state income tax data for 1999, which is the most recent state
data that is currently available. Table 10 shows a breakdown of state income taxes paid
by residents versus state income taxes paid by nonresidents. Table 11 shows only
information on all taxpayers but in more detail. The information in Table 10 was
necessary because nonresidents must report all income to the State of Iowa even though
they are required to pay taxes only on Iowa income. If this discrepancy is not considered,
the calculated average and marginal tax rates will be incorrect. In order to find the portion
of reported AGI and taxable income for which nonresidents must pay taxes, a series of
calculations were required. Definitions used are listed below.
22 / Lapan, Moschini, and Caruth
AGI = adjusted gross income (as defined by the Internal Revenue Service)
TI = taxable income
AGI per return = all taxpayer AGI/number of returns
Nonres AGI = all AGI – resident AGI
Resident TI = resident AGI * all taxpayer TI/all taxpayer AGI
Nonres TI = nonres AGI * all taxpayer TI/all taxpayer AGI
Nonres taxes paid = all taxpayer taxes paid – resident taxes paid
Number of nonres returns21 = AGI per return * nonres AGI
Number of resident returns = all taxpayer returns – number of nonres returns
Ave. resident TI = resident TI/number of resident returns
Ave. resident tax bill = resident tax paid/number of resident returns
Ave. marginal resident tax rate = statutory marginal rate for the calculated average TI
Ave. marginal nonres tax rate = ave. marginal resident tax rate
Ave. nonres tax bill = nonres tax paid/number of nonres returns
Ave. effective nonres TI = ave. nonres tax bill * ave. resident TI/ave. resident tax bill
Effective nonres TI = ave. effective nonres TI * number of nonres returns
Effective nonres AGI = effective nonres TI/nonres TI * nonres AGI
All taxpayer effective AGI = effective nonres AGI + resident AGI
All taxpayer effective TI = effective nonres TI + resident TI
Ave. tax rate = all taxpayer tax paid/all taxpayer effective TI
After income tax values had been calculated it became necessary to calculate the amount of
money each income group spent on the sales tax. Since sales tax returns are not reported to the
state in terms of individual payments, the consumer expenditure survey was used to estimate
individual payments.22 The survey gives information about the spending habits of people by
income groups. By examining each expense, and adding the amount spent on each taxable item,
it is possible to find the amount of money each income group spends on taxable items, as
illustrated in Table 12.23 It is then possible to find the amount paid as sales taxes by the
representative consumer in each group facing a 5 percent sales tax rate (as applies in Iowa), and
these estimates are reported in the last row of Table 12. Because Iowans are not exactly like the
individuals represented in the national survey, this figure must be modified to reflect the fact that
Are All Taxes Equally Bad? / 23
adjusted gross income within each group in Iowa differs slightly from its national counterpart
and also the fact that spending patterns here differ slightly. Specifically, for FY 1999, income
before taxes for midwesterners was 95.5 percent of average nationwide income before taxes, but
average expenditures for midwesterners was 98.1 percent of average nationwide expenditures,
indicating that midwesterners spend a greater percentage of their income.24 Thus, taxable
spending for each income cohort within Iowa was estimated by (i) multiplying the figure
obtained in Table 12 by 1.027 to reflect the greater average spending propensity of Iowans, and
(ii) multiplying this figure by the ratio of average Iowa income to average national income
within each group. Finally, the Iowa sales tax rate of 5 percent was applied to this spending
figure to estimate sales tax paid by each group. The total sales taxes imputed to Iowa by this
method differed from the total actual Iowa sales tax revenue by (only) 3.46 percent, so each
group’s estimated sales tax was scaled upward to reflect total sales tax revenue.25
For the lump-sum equivalent tax scenario, the sales tax expenditures for each group were
treated as though they were income tax payments. In order to find how much money would be
saved by itemizing, those payments were multiplied by the portion of taxpayers that itemized
their federal tax returns and the federal marginal tax rate that those people paid. With the
decrease in federal income taxes, Iowans were no longer able to deduct as much from their state
income taxes. It was then necessary to multiply the federal tax savings by the state marginal tax
rates in order to find the requisite increase in state taxes. This tax increase in turn was multiplied
by the federal marginal tax rates in order to find the decrease in federal taxes caused by the
second increase in state taxes. This process was iterated fifteen times in order to find a
reasonable approximation of the net savings to each income group.26 The result was the total
savings to Iowa taxpayers for 1999 if the entire sales tax was converted. This, however, resulted
in the state government revenue increasing by almost $8 million. In order to make the revenue
effects neutral, a small portion of the sales tax was eliminated but not converted to an incomebased tax. The revenue-neutral lump-sum equivalent tax scenario calculations were then repeated
taking this change into account. The calculations are shown in Table 13.
For the taxpayer-neutral lump-sum equivalent tax scenario, the entire sales tax was
converted to an income-based tax. Then, after all effects due to itemizing were calculated, the
state income tax was increased by the amount of the net savings from itemizing for each group.
Next, savings from itemizing were calculated for that tax increase. This process was completed
24 / Lapan, Moschini, and Caruth
fifteen times, so that the total savings from itemizing was totally counteracted by the income tax
increases. The sum of all savings from itemizing and incidental revenue increases is the total
revenue increase for the state government for 1999. The taxpayer-neutral lump-sum equivalent
tax scenario calculations are shown in Table 14. In order to adjust the dollar values from these
scenarios to 2002 dollars, the Consumer Price Index (CPI) was used. The average CPI for 2002
was divided by the average CPI for 1999 in order to obtain the inflation rate for the three-year
period. That inflation rate was multiplied by each of the 1999 savings in order to find the 2002
savings.27 The CPI data is shown in Table 15. For the flat tax equivalent scenarios, the first step
was to increase the average marginal tax rates and average tax rates in order to reflect the income
tax increase. After that was completed, the immediate increase in state income taxes was
calculated by multiplying the tax increase by the total effective taxable income for each group.
Finding the total money saved then required an iterative process similar to the one used in the
lump sum equivalent tax scenario. Table 16 shows the iterations for the scenario in which the
government breaks even. The results were then adjusted to 2002 values using the CPI data from
Table 15.
TABLE 9. Tax Year 1999, United States selected income and tax items for individual income tax returns: Forms 1040,
1040A & 1040EZ, (amounts are in thousands of dollars): State of Iowa, by Size of Adjusted Gross Income, Filing/
Processing Period: Jan 1, 2000 to Dec 31, 2000
Iowa
Returns count
Joint return count
Single return
count
Head of
household
AGI amount
No. of returns
Amount
Size of Adjusted Gross Income
$10,000
$20,000
Under
Under
$20,000
$30,000
Total
Returns
Breakeven
and Loss
$0.01
Under
$10,000
1,345,040
605,488
11,793
5,782
287,300
19,697
243,545
56,187
612,820
5,433
243,672
111,366
348
21,356
52,170,593
-572,703
1,433,422
381,460
1,791
6,742
5,846,823
29,463
149,755
$30,000
Under
$50,000
$50,000
Under
$75,000
$75,000
Under
$100,000
200,760
67,113
266,485
161,696
194,818
168,011
73,774
67,226
140,339
127,002
146,409
102,029
82,266
21,734
5,517
11,277
36,581
27,613
19,499
4,334
834
1,635
4,961,037
10,469,311
11,853,694
6,291,737
20,377,473
Total Itemized Deductions:
22,363
30,902
84,302
114,155
60,171
121,205
911,388
1,379,130
895,586
2,737,510
3,648,359
328,228
311,350
$75,000
and
Over
Taxable Income:
No. of returns
Amount
1,118,131
0
35,642,254
0
118,509
212,284
202,602
196,739
265,372
194,643
73,745
140,266
1,299,702
2,679,566
6,656,447
8,354,099
4,686,788
16,440,154
No. of returns
Amount
Average TI
Average tax bill
1,071,533
6,670,191
$26,499.03
41
360
$0.00
$4,959.10
$30.53
Ave tax rate
18.71%
% joint returns
45.54%
50.00%
114,859
30,412
$738.89
$105.85
181,394
174,342
$5,336.60
181,319
357,411
$13,347.11
259,319
917,544
$24,978.69
194,341
1,248,007
$42,881.56
73,732
853,652
$63,528.99
140,260
3,942,116
$117,146.01
$715.85
$1,780.29
$3,443.14
$6,406.01
$11,571.18
$28,089.95
14.33%
13.41%
13.34%
13.78%
14.94%
18.21%
23.98%
6.92%
23.49%
34.11%
61.37%
86.57%
91.37%
90.77%
Are All Taxes Equally Bad? / 25
Income Tax:
Iowa
Total
Returns
Breakeven
and Loss
$0.01
Under
$10,000
$10,000
Under
$20,000
$20,000
Under
$30,000
$30,000
Under
$50,000
$50,000
Under
$75,000
$75,000
Under
$100,000
75,000
and
Over
Joint return marg
rate(*)
15.00%
15.00%
15.00%
15.00%
15.00%
15.00%
15.00%
27.50%
30.50%
Single returns
46.09%
46.99%
85.58%
61.21%
51.86%
31.23%
11.20%
7.50%
8.06%
Single return
marg rate(*)
15.00%
15.00%
15.00%
15.00%
15.00%
15.00%
27.50%
27.50%
30.50%
% head of
household
Head of
household
marg rate(*)
8.38%
3.01%
7.50%
15.29%
14.03%
7.40%
2.23%
1.13%
1.17%
15.00%
15.00%
15.00%
15.00%
15.00%
15.00%
27.50%
27.50%
30.50%
Avg marg tax
rate
15.00%
15.00%
15.00%
15.00%
15.00%
15.00%
16.68%
27.50%
30.50%
Source: U.S. Internal Revenue Service, July 2001.
Note: (*) These rates are from the 2001 tax rate schedule.
26 / Lapan, Moschini, and Caruth
TABLE 9. Continued
Are all Taxes Equally Bad? / 27
TABLE 10. Tax Year 1999, Iowa income tax incidence for residents
All Taxpayers
Adjusted Income
Gross ($) Class ($)
No AGI
Resident Taxpayers
AGI ($)
Tax Paid ($)
AGI ($)
Tax Paid ($)
0
209,763
0
0
0
5,000
554,583,737
250,496
527,100,000
240,000
5,000
10,000
1,705,408,510
9,834,509
1,607,100,000
9,500,000
10,000
14,000
1,979,064,973
28,593,061
1,842,200,000
27,500,000
14,000
20,000
4,210,714,233
92,286,464
3,924,400,000
89,000,000
20,000
25,000
4,387,618,379
126,600,055
4,101,600,000
122,500,000
25,000
30,000
4,674,985,430
152,118,999
4,378,400,000
147,500,000
30,000
40,000
8,471,729,375
296,749,962
7,895,000,000
288,200,000
40,000
50,000
6,037,607,185
221,510,700
5,528,400,000
214,700,000
50,000
75,000
7,561,276,275
288,169,672
6,636,600,000
276,800,000
75,000
And over
29,295,367,654
655,423,752
12,345,300,000
603,100,000
Total
68,878,355,751
1,871,747,433 48,786,100,000 1,779,000,000
Source: Iowa Department of Revenue and Finance, July 2001.
28 / Lapan, Moschini, and Caruth
TABLE 11. Tax year 1999, Iowa selected income and tax items for individual income
tax returns by size of adjusted gross income
Adjusted Gross
Nt Taxable
Tax
Adjusted
Income Brackets No. of
Income ($)
Paid ($)
Returns Gross Income ($)
($)
NO
AGI
23768
0
2209332
209,763
1
5,000
192,255
554,583,737
287,322,127
250,496
5,000
10,000
229,367
1,705,408,510
1,192,660,066
9,834,509
10,000
14,000
165,042
1,979,064,973
1,475,893,017
28,593,061
14,000
20,000
247,840
4,210,714,233
3,248,976,370
92,286,464
20,000
25,000
195,396
4,387,618,379
3,444,705,849
126,600,055
25,000
30,000
170,462
4,674,985,430
3,669,118,128
152,118,999
30,000
40,000
244,998
8,471,729,375
6,555,529,016
296,749,962
40,000
50,000
135,770
6,037,607,185
4,573,329,652
221,510,700
50,000
75,000
And
Over
126,665
7,561,276,275
5,578,209,735
288,169,672
95,868
29,295,367,654
19,938,983,448
655,423,752
1,827,431
68,878,355,751
49,966,936,740
1,871,747,433
145,156
20,092,269,436
13,789,722,464
89,565,625
75,000
Total
Non-Resident
Total
Source: Iowa Department of Revenue and Finance, July 2001.
TABLE 12. Consumer expenditure survey data (U.S. national data, year 1999)
Item
Number of consumer units (in
thousands)
Income before taxes ($)
Income after taxes ($)
Average annual expenditures
($)
Food away from home ($)
Alcoholic beverages ($)
Maint, repairs, ins, other ($)
Estimated taxable spending ($)
$10,000 to
$20,000
$20,000 to
$30,000
$30,000 to
$40,000
$40,000 to
$50,000
$50,000 to
$70,000
$70,000
and Over
11,497
5,592
5,420
15,634
14,563
14,295
11,560
24,467
23,487
9,453
34,353
32,458
7,381
44,321
41,405
10,999
58,473
54,073
15,168
113,441
102,616
15,962
864
158
358
21,794
1,092
191
584
28,916
1,625
267
768
35,048
2,142
292
748
40,826
2,365
345
766
49,606
2,803
443
1,002
76,742
4,398
696
1,869
1,440
130
251
1,878
190
313
2,159
213
451
2,298
207
515
2,491
271
575
2,795
359
784
3,412
965
945
560
799
0
790
84
285
279
733
821
1,100
6
1,280
128
494
473
989
1,127
1,553
14
1,781
237
576
677
1,323
1,343
1,904
86
2,296
360
705
798
1,681
1,549
1,677
43
2,610
414
811
870
1,882
2,188
2,139
33
3,145
533
870
1,050
2,754
3,431
3,625
75
4,322
585
1,200
1,308
4,121
224
68
527
247
104
262
385
132
309
452
147
347
500
166
425
525
209
602
794
330
1,430
235
274
305
336
376
391
328
7,785.28
10,425.48
13,902.00
16,657.00
18,136.00
22,625.00
33,834.00
Source: U.S. Bureau of Labor Statistics, May 2001.
Are All Taxes Equally Bad? / 29
Utilities, fuels, and public
services ($)
Other household expenses ($)
Housekeeping supplies ($)
Household furnishings and
equipment ($)
Apparel and services ($)
Other vehicles ($)
Other vehicle expenses ($)
Vehicle finance charges ($)
Maintenance and repairs ($)
Vehicle insurance ($)
Entertainment ($)
Personal care products and
services ($)
Reading ($)
Education ($)
Tobacco products and
smoking supplies ($)
Less Than
$10,000
AGI Bracket
Money to sales
Tax ($)
Part of sales tax
not replaced by
income tax ($)
Decrease in
federal taxes 1
($)
Increase in state
taxes 1 ($)
Decrease in
federal taxes 2
($)
Increase in state
taxes 2 ($)
Decrease in
federal taxes 3
($)
Net savings to
taxpayers ($)
Less than
$10,000
$10,000 to
$20,000
163,848,718
227,950,218
3,691
$20,000 to
$30,000
$50,000 to
$70,000
$70,000
and Over
1999
Total
Total Adjusted
to 2002
$30,000 to
$40,000
$40,000 to
$50,000
264,650,679
209,715,959
124,177,986
109,247,104
126,526
349,510
617,156
379,270
784,364
4,820,571
7,081,088
7,620,681
576,737
3,137,914
6,102,387
9,922,199
5,874,522
10,600,230
61,510,744
97,724,732
105,171,550
21,641
176,145
384,796
674,710
399,467
839,538
5,448,909
7,945,206
8,550,647
3,246
26,422
57,719
101,206
59,920
140,026
1,661,917
2,050,457
2,206,706
122
1,483
3,640
6,882
4,075
11,090
147,220
174,512
187,810
18
222
546
1,032
611
1,850
44,902
49,182
52,930
561,928
3,113,444
6,121,698
9,959,942
5,910,745
10,675,718
62,439,165
98,782,640
106,310,073
238,332,453 1,337,923,119
1,429,402,113
30 / Lapan, Moschini, and Caruth
TABLE 13. Net savings due to itemizing the revenue-neutral lump-sum equivalent tax
TABLE 14. Total tax increase by income group for taxpayer-neutral lump-sum equivalent tax
AGI Bracket
Savings from
itemizing initial
($)
Incidental Increase
in tax revenue 1
($)
Savings from
itemizing 1 ($)
Incidental Increase
in tax revenue 2
($)
Savings from
itemizing 2 ($)
Total ($)
Less than
$10,000
$10,000 to
$20,000
$20,000 to
$30,000
$30,000 to
$40,000
$40,000 to
$50,000
$50,000 to
$70,000
$70,000
and Over
1999
TOTAL
2002 Adjusted
Total
558,250
2,988,578
5,779,821
9,370,361
5,548,421
9,962,884
58,808,060
93,016,375
100,104,407
21,764
177,740
388,984
683,674
404,820
856,929
5,715,827
8,249,739
8,878,385
1,902
39,182
126,228
418,679
247,910
908,574
14,510,772
16,253,247
17,491,777
74
2,330
8,495
30,547
18,088
78,148
1,410,369
1,548,052
1,666,017
6
514
2,757
18,707
11,077
82,858
3,580,504
3,696,423
3,978,098
581,997
3,208,381
6,306,535
10,524,272
6,231,681
11,905,551
85,660,427
124,418,845
133,899,808
TABLE 15. Consumer Price Index (Midwest urban)
Year
CPI
162.7
168.3
172.8
175.1*
Source: U.S. Bureau of Labor Statistics, June 2002.
* Estimate for 2002 is based on a linear regression using monthly data for the period 1987-2002.
Are All Taxes Equally Bad? / 31
1999
2000
2001
2002
Adjusted Gross
Income Class
No AGI
$0 to $5,000
$5,000 to
$10,000
$10,000 to
$14,000
$14,000 to
$20,000
$20,000 to
$25,000
$25,000 to
$30,000
$30,000 to
$40,000
$40,000 to
$50,000
$50,000 to
$75,000
$75,000 and
over
Total
Immediate
Money Saved
Increase in
Itemizing
Income Taxes
Federal
($)
Taxes 1 ($)
77,432
1,764
Change in
IncomeTax
Revenue After
Itemizing 1 ($)
77,500
Money Saved
Itemizing
Federal Taxes 2
($)
1,774
Change in
Income Tax
Revenue After
Itemizing 2 ($)
77,500
Money Saved
Money Saved
Itemizing
Itemizing
Federal Taxes 15
Federal Taxes
($)
Adjusted to 2002 ($)
1,774
1,909.42
9,957,551
35,051
9,959,032
35,273
9,959,041
35,274
37,962.22
40,777,356
143,537
40,788,846
145,260
40,788,984
145,281
156,352.00
50,063,176
689,542
50,118,372
697,821
50,119,035
697,922
751,104.71
110,045,145
1,515,699
110,191,027
1,537,581
110,193,133
1,537,902
1,655,093.20
116,636,167
2,692,985
116,895,361
2,731,864
116,899,103
2,732,433
2,940,650.18
124,207,306
2,867,793
124,493,648
2,910,744
124,497,937
2,911,397
3,133,251.70
220,466,869
10,461,638
221,544,917
10,623,345
221,561,580
10,625,884
11,435,597.71
151,421,872
7,185,301
152,162,301
7,296,365
152,173,746
7,298,109
7,854,239.60
178,643,162
17,459,070
180,637,819
17,791,758
180,675,828
17,798,221
19,154,480.36
320,034,937
84,302,293
330,559,879
87,512,401
330,960,654
87,639,476
94,317,777.85
1,322,330,973
127,354,673
1,337,428,702
131,284,188
1,337,906,541
131,423,674
141,438,418.96
32 / Lapan, Moschini, and Caruth
TABLE 16. Change in income tax revenue and money saved itemizing for the revenue-neutral flat tax equivalent scenario
Endnotes
1. For example, if a sales tax brings the total price of a good above a consumer’s
willingness to pay, the consumer will choose not to consume the taxed product. In that
case, the government does not collect anything while the consumer is worse off because
he or she is forced to allocate income to a less desirable consumption bundle. The only
tax believed to be nondistortionary is a lump-sum tax (i.e., a fixed amount levied on an
individual irrespective of any economic decision he or she makes), but differential
lump-sum taxes also are widely considered infeasible on grounds of equity.
2. Contributions in this area include Ramsey 1927, Mirrlees 1971, Diamond and
Mirrlees 1971, and Atkinson and Stiglitz 1976. For a textbook presentation of
alternative models and results, see Myles 1995.
3. The logic for this deduction, it seems, is that taxes paid to the state government are
not discretionary and therefore are not available as income to the taxpayer.
4. In general, a tax schedule is said to be “progressive” or “regressive” according to
whether the tax burden is borne relatively more by the wealthy or by the poor. More
specifically, a tax schedule is said to be “progressive” if the average tax rate increases
as taxable income rises and regressive otherwise.
5. For a full list of services that are subject to the Iowa sales tax go to
http://www.state.ia.us/tax/educate/78539.html.
6. For a complete list of cities in Iowa and their respective local option sales taxes, go to
http://www.state.ia.us/tax/business/lostcit.pdf.
7. This figure represents tax return data for consistency with the rest of the analysis,
although it is recognized that a distribution of households’ income may be somewhat
more informative here.
8. For more information regarding the distribution of state tax burdens, see Ettlinger et
al. 1996.
9. Missouri also allows a federal income tax deduction, but unlike Iowa’s, it is not
unlimited. Missouri law allows for a maximum federal income tax deduction of
$5,000 for single returns and $10,000 for joint returns.
10. The sales tax data for Figures 5 and 6 include local option sales tax revenue that is
collected by the state and then redistributed to the local municipality. The local option
sales taxes are not, however, included in later analyses because they constitute a local
34 / Lapan, Moschini, and Caruth
tax instead of a state tax. Local option sales taxes generally account for approximately
10 percent of gross sales tax collections.
11. For a more thorough discussion on trends in state taxes, see the Center for Budget and
Policy Priorities 2002. For more information on the distributional effects of Iowa’s
sales tax increase and income tax decrease, see Fisher and Bruner 2002.
12. The state sales tax deduction was eliminated from federal law in 1986 as part of an
effort to simplify federal income taxes. Lawmakers eliminated a variety of deductions
and replaced them with greater personal exemptions and a larger standard deduction.
In April 2002, Congressman Bryant of Tennessee introduced a bill to allow taxpayers
the option to deduct either sales or income taxes from their federal tax returns. For
more information on the bill, see Bryant 2002.
13. This value is based on the authors’ calculations using state and federal income tax
data. A similar result can be found in Slocum et al. 1998.
14. There is actually an additional economic reason suggesting that the taxpayer could in
fact benefit from a lump-sum tax replacing the sales tax, even without the considerations related to the federal deduction. Because currently some goods and services are
not taxed, eliminating the sales taxes would in fact affect relative prices, and the
taxpayer/consumer could actually benefit by changing their purchasing decisions.
Although this argument does strengthen our point here, in this study we ignore such
second-order effects because they are likely to be small relative to the main impact
that we are interested in quantifying.
15. As the individual’s federal tax liabilities are reduced, the individual’s income that is
taxable from the State of Iowa’s perspective increases, which would, ceteris paribus,
actually result in higher revenues than Iowa’s current income tax structure provides.
In all calculations subsequently presented, we account for such induced effects.
16. The most recent available data was for tax year 1999. These data were used to
calculate a net benefit of $98.8 million for this scenario. The value was then adjusted
to 2002 dollars using the consumer price index. This inflation adjustment will tend to
underestimate the actual value because it does not account for real growth in the state
of Iowa.
17. The fact that such gains would constitute a flow that Iowa taxpayers would enjoy
indefinitely should also be stressed. Because of that, the present-value concept (i.e.,
the discounted accumulation of all future gains in this flow) would perhaps be a more
appropriate criterion for assessing the value of the policy change being advocated.
For example, with a 5 percent discount rate, the present value of all future net benefits
arising from replacing the sales tax is in excess of $2 billion!
18. The logic for this was discussed earlier, and here it works as follows. Were the
government to set the sales replacement tax rate at 3.85 percent, it would raise exactly
the amount captured by taxpayers in the preceding scenario ($141.4 million
Are all Taxes Equally Bad? / 35
annually). But in such a case, the increase in state income tax would allow itemizing
taxpayers to decrease their federal tax burdens. Accordingly, for the Iowa government
to leave the total tax burden unchanged, the Iowa tax rate can be increased further,
specifically to 3.89 percent, such that the total additional revenue obtained by the
state is estimated to be $157 million annually.
19. A list of Iowa sales tax exemptions, and the dates that they were enacted, can be
found at http://www.state.ia.us/tax/educate/79120.html.
20. Iowa government spending information can be found at
http://staffweb.legis.state.ia.us/lfb/factbook/Iowa_Factbook_2001.pdf or
http://staffweb.legis.state.ia.us/lfb/miscpubs/ladar_fy2003.pdf
21. The given equation was not adequate for the lowest income bracket or the highest
income bracket. Those brackets were calculated separately. For the lowest income
bracket, the equation was number of nonres returns = all taxpayer returns * total
nonres returns/total all taxpayer returns. For the highest income bracket, the equation
was number of nonres returns = total nonres returns – sum of other brackets’ nonres
returns.
22. A similar procedure was used by Swenson (1999).
23. The information in Table 12 can be found at
ftp://ftp.bls.gov/pub/special.requests/ce/standard/1999/income.txt.
24. Because midwesterners spend 98.1 percent of the national average, while their
income is only 95.5 percent of the national average, the average midwesterner spends
about 2.7 percent more of their income (98.1/95.5=1.027). Information on regional
income and expenditures can be found at
ftp://ftp.bls.gov/pub/special.requests/ce/standard/1999/region.txt.
25. The calculated sales tax revenue was $1.29 billion whereas the actual revenue was
$1.34 billion. There could be a variety of reasons for this discrepancy, including
differences in spending patterns between the Midwest and Iowa, unreported income,
or spending by out-of-state visitors.
26. The iterations could be continued indefinitely, but changes in the final result would be
less than $0.01.
27. Although this procedure takes inflation into account, it does not include any estimate
of real growth. Consequently, these adjustments should tend to underestimate the
difference between 1999 values and 2002 values.
References
Atkinson, A.B., and Stiglitz, J.E. 1976. “The Design of Tax Structure: Direct Versus Indirect Taxation.”
Journal of Public Economics 6: 55-75.
Bruce, D., and W.F. Fox. State and Local Sales Tax Revenue Losses from E-Commerce: Updated
Estimates. Center for Business and Economic Research, The University of Tennessee, September.
http://www.statestudies.org/ecomreport.pdf (accessed August 2002).
Bryant, E. 2002. “Sales Tax Fairness Act of 2002.” H.R. 4487. April.
http://www.house.gov/bryant/hr4487.pdf (accessed August 2002).
Center on Budget and Policy Priorities. 2002. State Tax Systems Are Becoming Increasingly Inequitable.
January. http://www.centeronbudget.org/1-15-02sfp-pr.htm (accessed August 2002).
Diamond, P.A., and J.A. Mirrlees. 1971. “Optimal Taxation and Public Production.” Part 2: “Tax Rules.”
American Economic Review 61: 261-78.
Economist. 2002. “State Budgets—Tangled Up in Red.” 13 July, pp. 25-26.
Ettlinger, M.P., R.S. McIntyre, E.A. Fray, J.F. O’Hare, J. King, and N. Miransky. 1996. Who Pays? A
Distributional Analysis of the Tax Systems in All Fifty States. Citizens for Tax Justice. June.
http://www.ctj.org/html/whopay.htm (accessed August 2002).
Feldstein, M. 1976. “On the Theory of Tax Reform.” Journal of Public Economics 6: 77-104.
Fisher, P., and C. Bruner. 2002. Tax Policy and Economic Growth in Iowa: Who Gained in the 1990s?
Mount Vernon, IA: The Iowa Policy Project, January. http://www.iowapolicyproject.org/020122polbrf-taxeq.pdf (accessed August 2002).
Iowa Department of Revenue and Finance. 2001a. 1999 Iowa Individual Income Tax Annual Statistical
Report. Des Moines, IA. July. http://www.state.ia.us/tax/educate/99increp.pdf (accessed August
2002).
———. 2001b. Annual Report for Fiscal Year 2001 Iowa Department of Revenue and Finance. Des
Moines, IA. 30 November. http://www.state.ia.us/tax/educate/0178508.pdf (accessed August 2002).
———. 2000. Iowa Sales and Use Tax Annual Report for Fiscal Year 1999. Des Moines, IA. April.
http://www.state.ia.us/tax/educate/SUTA99.pdf (accessed August 2002).
Mirrlees, J.A. 1971. “An Exploration in the Theory of Optimum Income Taxation.” Review of Economic
Studies 38: 175-208.
Musgrave, R.A. 1985. “A Brief History of Fiscal Doctrine.” In Handbook of Public Economics, vol 1.
Edited by A.J. Auerbach and M. Feldstein. Amsterdam: Elsevier Science Publishers.
Myles, G.D. 1995. Public Economics. Cambridge: Cambridge University Press.
Are all Taxes Equally Bad? / 37
Ramsey, F.P. 1927. “A Contribution to the Theory of Taxation.” Economic Journal 37: 47-61.
Slocum, T., R.G. Lynch, M.P. Ettlinger, and M.W. Gardner. 1998. Choices for Iowa: Building A Better Tax
System. Washington, D.C.: Institute on Taxation and Economic Policy. September.
http://www.ctj.org/html/iamenu.htm (accessed August 2002).
Swenson, D. 1999. “Equity, Expediency, and the Local Option Sales Tax.” Unpublished manuscript,
Department of Economics, Iowa State University, October.
U.S. Bureau of Labor Statistics. 2001. Consumer Expenditures in 1999. Report 949. Washington, D.C.
May. http://www.bls.gov/cex/csxann99.pdf (accessed August 2002).
———. 2002. Consumer Price Index Home Page. Washington, D.C. June.
http://www.bls.gov/cpi/home.htm (accessed August 2002).
U.S. Census Bureau. 2002. American Factfinder. Washington, D.C. July.
http://factfinder.census.gov/servlet/BasicFactsServlet (accessed August 2002).
U.S. Internal Revenue Service. 2001. Tax Year 1999, United States Selected Income and Tax Items for
Individual Income Tax Returns [spreadsheet online]. Washington, D.C. July.
http://www.irs.gov/pub/irs-soi/99in16ia.xls (accessed August 2002).